Defining the optimal cardiac troponin T threshold for predicting death caused by periprocedural myocardial infarction after percutaneous coronary intervention

Background-There is controversy about the diagnostic and prognostic significance of percutaneous coronary intervention- related myocardial infarction, especially with the use of cardiac troponin T (cTnT). This analysis was designed to address the question of the presence and the level of a prognostic cTnT threshold. Methods and Results-We evaluated 5268 consecutive patients who underwent nonemergent percutaneous coronary intervention between 2000 and 2009 with a preprocedural cTnT level below the upper limit of normal (ULN, ≤0.01 ng/mL). Postprocedural cTnT and creatine kinase-MB mass levels (ULN, 6.7 ng/mL in men and 3.8 ng/mL in women) were found to be associated with 3-month mortality in Cox proportional hazard models (hazard ratio per doubling of cTnT, 1.24; 95% confidence interval, 1.08-1.43; P=0.003 and hazard ratio per doubling of creatine kinase-MB, 1.30; 95% confidence interval, 1.05-1.60; P=0.018), adjusted for the Mayo Clinic risk scores for in-hospital and postdischarge mortality. The optimal prognostic threshold for 3-month mortality was 25× ULN for cTnT (hazard ratio, 4.53; 99% confidence interval, 1.59-12.9; P<0.001), which provided similar information as a value of 5× ULN for creatine kinase-MB (hazard ratio, 4.31; 99% confidence interval, 1.27-14.6; P=0.002). The cumulative mortality rate was 0.6% at 91 days. Conclusions-A significant association of postpercutaneous coronary intervention cardiac biomarker elevation with a small number of postpercutaneous coronary intervention outcomes was noted for the early (first 91 days) follow-up period with an identifiable optimal threshold of 25× ULN (0.25, ng/mL) for cTnT, which provided similar early outcome information as a cutoff of 5× ULN for creatine kinase-MB.